Blog runs on a public Sanity dataset (project id ships in their cdn.sanity.io image URLs); the site itself is a client-rendered SPA with no feed and no sitemap entries. Engineering + AI categories, 70 posts, active (Aug 2026). The old 'ShareChat TechByte' Medium publication (dormant since 2022) stays in the corpus under this source id.
ShareChat AI team has embarked on a journey of conducting cutting-edge AI research and building AI-powered solutions for providing best-in-class and safe user experience to our users. Our endeavour is to build solutions that can scale across diverse geographies and cultures…
ShareChat AI team made its presence felt at the recently concluded EMNLP Conference 2021 held in the Dominican Republic. Here’s what we spoke about!
In this approach, an additional calibration step is involved, wherein a representative dataset is used to estimate the range of activations using the variations in the dataset.
Written by Akash Manna, Vikram Gupta, Debdoot Mukherjee Every day, ShareChat and Moj receive millions of User Generated Content (UGC) pieces. To derive insights from these content pieces and recommend relevant and interesting content to our users, we require accurate, fast and…
Written by Sumit Malpani, Sumit Agrawal, Sidharth Mishra, Nitesh Jain “Creator Economy”, led by the rise of short-form videos, has become a global phenomenon in recent years. Today, there are more than 100M+ content creators globally, with increasing numbers of short video first…
We wanted to make the whole gifting experience feel very fast but this was not possible if we kept waiting for the API response.
Written by Gulzar Ahmed, Chetan Potnuru We recently shipped out a new Virtual Gifting experience for our ShareChat’s Audio Chatroom- a core feature where the audience can support the creators by sending them virtual gifts. In this blog, we’ll talk about why and how we revamped…
Written by Rishabh Jain, Praveen Dhinwa, Vikram Gupta, Hastagiri Vanchinathan, Debdoot Mukherjee ShareChat is India’s leading social media platform & Moj is India’s number one short video platform with one of the highest monthly active users. Both allow users to share their…
Though our problem from an outside perspective is similar to OCR in the wild problem, the primary distinction is content intelligence.
Written by Tejas Singh, Devyani Jain, Shivani Nayaka, Rishabh Dsouza and Anish Pratik Yayavaram One of the questions that we are often asked is, ‘Who is a ShareChat creator?’ And to be honest, it would be pretty hard to answer this question in a sentence or two. Rather, it would…
While designing the OCR system, we had both computational complexity and correctness in mind, as the system needs to scale up to the heavy usage and use-cases. We tried models, like Single Shot Detection[4] and R2CNN[5],
Majority of existing work tackles only one of the aforementioned problems at a time. We at ShareChat and Moj, on the other hand, are working towards a robust all-in-one framework to considerably improve the quality of the UGC,
Since our use cases deal primarily with spoken audio, we use the popular overlap and shift paradigm used to analyse speech signals with a 25ms window length and 10ms shift.
For each language and content type, we have a different configuration for extracting features as well as separate Milvus clusters. This allows us to tackle challenges that come with each language/content-type
A PID controller continuously calculates an error value as the difference between the desired value and the measured value and applies a correction factor so that the measured value becomes equal to the desired value.
The jobs are written in a way that most of the heavy lifting in terms of computation is done by BigQuery itself (parsed from GKE metering). We simply aggregate the results and push those back to another BigQuery dataset.
Millions of posts are created every day on ShareChat and Moj which makes timely and precise detection of Integrity Violating Content (IVC) crucial for ensuring the safety of our users and
If you have troubleshot Distributed Systems, you would realise that the maximum time is spent in trying to find out where the problem emanates from. Once the system-at-fault has been identified, the fix itself becomes very straightforward.
We used philosophies of open tracing for extracting various metrics from every single resource and collected Externally to create actionable items and build more intelligent infrastructure.
While our models demonstrate amazing speed and accuracy, they are data-hungry and require many labelled data for training. Labelling this data is a time consuming and costly process as it requires manual inspection.
Videos are an expressive, engaging and entertaining form of content. However, from a machine learning point of view, videos are incredibly complex as the information flows in both spatial and temporal dimensions.
Millions of posts are created every day on ShareChat and Moj which makes timely and precise detection of Integrity Violating Content (IVC) crucial for ensuring the safety of our users and the integrity of our platform.
Elasticsearch is a distributed document store that stores data in an inverted index. An inverted index lists every unique word that appears in any document and identifies all of the documents each
We just went through different solutions that we could use for the problem we have on our hands. All of these are valid solutions that could be used for different levels of scale.
our code needs to compare the number of times a user has already seen an ad to the configured frequency in real-time and for that, we need storage of how many times a user has seen an ad. On the face of it.
Furthermore, we have extensively used the bundle analysis tool for webpack called Bundle Analyser, which enables us to visualize both uncompressed and gzipped size of webpack output files with an interactive zoomable tree-map.
Automatic Natural Language Generation (NLG) has had a long tradition in the Computational Linguistics community, but the recent advances in deep learning have given it a new impetus. Its importance in the modern world cannot be overstated.
This article talks about some interesting scenarios we encountered with BT and the troubleshooting steps we followed through the entire process.
Netflix Archaius provides a DynamoDB client library with a FixedDelayPollingScheduler, which basically polls the config in fixed intervals. This was integrated to fetch sync/async thresholds, ignored APIs and similar other configs.
With an intent to make all deployments fully CD-compatible, the Infra team at ShareChat started to experiment with Spinnaker and we are happy to share our journey with you.
For a query post, inverted indices give us the potential result candidates for ranking. We then retrieve the codebooks for candidates, and also quantize the query to construct its codebook, according to its sub-vectors.
At ShareChat, our CNN architecture is based on DenseNet1, or ‘dense convolutional network’. DenseNet has a feed-forward processing layer architecture, connecting each layer with every other layer.